Verifiability, CT and Socionics

Index Forums Cognitive Functions Verifiability, CT and Socionics

  • Auburn
    Keymaster
    • Type: TiNe
    • Development: l--l
    • Attitude: Adaptive

    Importing from Discord for more dedicated discussion

     



     

    Auburn | TiNe

    great little talk above on Socionics vs CT @Echidna1000 @Mysti | Arcturian . i’d like to migrate over something i wrote to Ben earlier:

    mmh. well, just a heads up, after reviewing what i managed of the video (it may have been 1/3 since i also skipped to the end and watched the end) CT is …almost irreconcilably different than socionics’ models A and G. it would appear that the only parallels the two models have to CT is in some shared nomenclature but the semantic definitions behind the terms are quite at odds the way CT is built is with a physical (materialistic) analog to each postulated category of differentiation — so that there are no metaphysics at the core. in this sense, every postulated dichotomy is in constant testing against empirical data this also heavily constraints the number of categories that can be conjured up, and in an ideal scenario nothing is added to the model which cannot be traced back to that vultological source, and contested on those grounds thus– CT is also far more sparse when it comes to ‘theory’, as it doesn’t have the luxury of free-reign speculation

    
    

    Auburn | TiNe

    the way CT is phrased is around a naturally occurring phenomenon, and is a description of that natural phenomenon which causes the data to organize along said constraints. So it’s a data-before-theory approach, and that has a great methodological advantage. This brings me first to Mysti’s question (i’ll get to Jack’s):

    There’s long pages of explanations on the website but I haven’t exactly found exactly how the signals were determined as correlated to functions.

    
    

    One thing about CT is that it doesn’t take the existence of the functions for granted. First, it must be understood that the default position in academia right now is that the functions aren’t real. There has been nobody that’s confirmed their reality outside of the logic of self consistent systems, essentially rendering functions as concepts for coherent information organization. But CT approaches type differently. The question of type’s reality or the function’s reality isn’t measured against MBTI or Socionics, but instead against a different field. So for example, CT can be tested against neuroscience or genetics but not against socionics as the baseline of comparison because socionics is an inadequate baseline of comparison until and unless it has already proven its own functions to be objectively real. Thus, the playing field right now, as it stands in the academic body, is that no jungian typing system has provided experimental support for the postulated functions, in any form. The 4-letter dichotomies even have a little more methodological support, due to their proximity to Big Five, but even they are ephemerally supported.

    It’s really quite a severe situation. We may squabble over this or that typology system internally, among the community, but we’re all up a creek.
    Anyhow,
    Having said that,

    Echidna1000Yesterday at 10:45 PM objective verification is a tricky thing for a visual identification methodology. Establishing test-retest reliability is only the first part of the battle. The second is criterion validity. Pod’Lair had pretty good test-retest reliability, but they never got that criterion validity so they ended up smudging up anything concrete about their theory and it ceased to be a personality theory, just a theory of facial and manual movements.

    
    

    Yep, this is the right question. So, as far as test-retest reliability goes, I’ve been working on that the past few months by building the most airtight and qualia-eliminated set of definitions to date for the CT signals. This is because test-retest reliability is contingent on very clear definitions. Here is where I am in my progress atm: https://cognitivetype.com/test-codifier/

    ^ If you click the “+” icons next to the signal names (in those that have it) you’ll see several animated GIFs demonstrating the signal in question, as well as a quick description of precisely what is meant to be seen.

    So far we’ve been running tests and the last test was last month, where a considerable amount of members got 80%-85% consensus in readings.
    But I hope that with a standardization at this level, we can reach 95%+ in the next quiz, which I’m hoping takes place as soon as I finish the GIFs for all the signals.

    The second is criterion validity

    Yes, it aggravates me that someone like Dave Powers has repeatability but I disagree quite powerfully on what the value is of what he is measuring. So I’m glad you bring this up.

    Now, I haven’t yet run the stats but CT has a database of samples here: https://cognitivetype.com/database/ – each sample has a dedicated 2-2:30 minute video provided. So far the database is at ~580 samples or so. (we’ve typed way more, but it takes a lot of effort to clean up samples and provide video for them)
    Anyone can easily run data analyses on this database, even non-CT related people.
    I believe if they did so [and we’ll do so too, soon] they’d see quite a lot of solid patterns rise up; clearly in the category of statistically significant. For example, the correlation of CT’s Je-leads with politics, life coaching, logistics, leadership is very strong just by grouping these people together vultologically. Far above average.
    Now if indeed CT extracted the profile of a “Je” type from the data, then this should be reflected in our database by looking at all the Je samples.

    But equally, CT predicts that if another group wanted to run an independent experiment (in a different country, lets say) and look for Je-leads using the vultology code, that the same patterns would re-appear, which are listed in the profile of Je. The Je profile would be re-written from new information.

    Pod’Lair had pretty good test-retest reliability, but they never got that criterion validity

    Indeed. The difference between CT and Podlair is that CT is actually quantifying its signals in the form of GIF catalogs, and also providing a database of videos so that anyone can work out the typing for themselves by reference to the GIF catalogs, and/or learn the reading methodology for themselves and see if it scales or doesn’t scale elsewhere. But also, they can see whether or not it corresponds to the thing being mentioned, and again, a holistic analysis of everyone within the given visual group should reproduce the same profiles over and over.

    Granted, CT may be wrong, but it puts forth all the tools anyone would need, to debunk it.

    Or to verify it.
    (well i should say, i’ve been struggling to put those tools forward, moving out of R&D, in the past year. i spent the past year building up the database for this purpose; to have testability)
    (and now i’m working on the GIF catalogs for the same reason)
    • This topic was modified 1 year, 3 months ago by Auburn.
    • This topic was modified 1 year, 3 months ago by Auburn.
    Auburn
    Keymaster
    • Type: TiNe
    • Development: l--l
    • Attitude: Adaptive

    The Doctor | Pleiadian

    The primary flaw I see in this is similar to the Pod’Lair and OP issues. The reason they’re getting solid test-retest reliability is because they’ve established clear expectations for subjective interpretations. While their subjective interpretations are more abstract than CT’s, CT still relies upon the assumption that signal X corresponds to function A – which could be incorrect – and the interpretation of that signal is still subjective. This is in effect self validating, but relying on empirical factors rather than abstract factors, which has an inherent advantage over the other approaches. Also, while Nardi hasn’t proven that cognitive functions exist to a degree that has been accepted by the academic community, his coherence map data set is quite compelling, with an r = ~.9 within type. I’d love to get you two connected and comparing data, as you’re both focused on physical manifestations. As for Socionics, I’m very interested in what degree any given person’s CT function signals might manifest in accordance with their Socionics model – especially with respect to the ‘strong but unvalued’ functions. I am seeing a fair amount of signals that imply this might be the case.

    The Doctor | Pleiadian

    podlair is such a good example of what not to do tho.

    OP is a great example of training people to agree with you to create a high test-retest accuracy.

    Dave and Shannon have many times talked about how long it took Shannon to be able to accurately predict the same way as Dave, and the process they described was very clearly Dave telling her she was wrong until she learned how to look at things the same way he did. That only proves that Shannon is good at predicting Dave, not that their system is accurate.

    The Doctor
    Participant
    • Type: Unknown
    • Development:
    • Attitude: Unknown

    In response to a question from:

    Crystal Waterfall🍸Today at 2:19 AM
    I know you have good intentions…but what do you mean by another level of verification?

    What I mean is that while cognitive function theory wasn’t invented by CT, CT has its own definitions of functions, even if they overlap strongly with Jung’s original work. Because those definitions (even Jung’s) are not verifiable, they are subjective.

    CT also created its own interpretations of micro expressions and body language. This can only be verified by subjective interpretation (at current). Therefore, both of these methods are subjective – even if they’re accurate.

    What I am proposing is a third means of verification that is external to CT and objective. Nardi’s EEG scans might be a good choice as they don’t attempt to measure either method – personality traits or interpretation of body language.

    But, any objective means that could provide solid correlation would work. The reason this is important is that it removes the counter argument that CT is inherently subjective, just with two steps, one of them including empirical interpretation of body language, which isn’t normally associated with personality. Removing this argument would go a long way toward setting up CT as a verifiable system – even if it’s something as simple as CT types people the same as Nardi’s scans 90% of the time.

    (Note: I’m only going back to Nardi’s scans so often because I’m familiar with them, and I have no other ideas on what else could work… if I did I’d have suggested it.)

    • This reply was modified 1 year, 3 months ago by The Doctor.
    • This reply was modified 1 year, 3 months ago by The Doctor.
    Auburn
    Keymaster
    • Type: TiNe
    • Development: l--l
    • Attitude: Adaptive

    CT still relies upon the assumption that signal X corresponds to function A – which could be incorrect – and the interpretation of that signal is still subjective.

    well, … not quite. because *[function A]* doesn’t exist as far as science is concerned. none of the functions are “real” according to them, and that is the appropriate position to take until further notice. So there is no point of comparison to say they’re correct or incorrect in saying so and so is or isn’t a function.

    At which point, we must ask.. if the functions are real at all, what proof do we have of them?

    And just what is their reality?

    In what way are they real and what do they mean?

    This is often the skipped-over aspect of most typologies, who take for granted Jung’s testimony. The truth is Jung is an insufficient authority on the matter, and any reality ‘functions’ might have would have to be seen, and exist as a verifiable phenomenon independent of Jung’s conjectures.

    And then that phenomenon would be seen to have a correlate in something an old Swiss psychologist first postulated in rougher language and with less methodological constraints.

    The best typology could hope for is for there to be a genuine scientific discovery of some sort, that confirms what they have been thinking is true by proxy of its proximity to the original conjecture. But even in such a case, it would not come in the form that typologist have been thinking. Certainly not in full– as any translation from conjecture to reality alters the outcome, as its understood in its sharper-resolution reality.

    If the’s any reality to the functions, it has to be independently generated out of evidence that is atheoretically derivable, else typology remains metaphysical. If one wishes to keep their typology in a rationalistic dome, that’s fine.

    The proper way to look at CT

    Now, what CT is putting forth is evidence for the existence of [a phenomenon (what is it? we are still seeing)] which is self-existent. In the strictest of sense it is inappropriate to call them by the 8 function names, and if a scientific body was aiming to verify CT they’d give these signal clusters (with a high Cronbach’s alpha value) a technical name like VC-31.

    We then might find that VC-31 (lets say we nickname it Ne) and XZ-53 (lets say we nickname it Si) are strongly correlated with each other too, and both VC-31 and XZ-53 are found statistically in opposition to UU-32 and UX-43 (Se/Ni).

    If such is seen as true, then we have evidence for a non-Jungian phenomenon that is simply a set of statistically significant, double-humped (bimodal distribution) values:

    ^ An uneven distribution like this can be tested, by seeing/clicking the signals in the CT code. We can check how often and how strong signal mixing is, and determine organically whether there are certain visual traits in the population that show differentiation of this sort.

    If indeed they do, then we have an objective metric for the existence for something like M (“Ne/Si”) vs V (“Ni/Se”), as visual categories. Now, to be absolutely fair it would stay at just that, until further notice. Even if there was statistically significant exclusion between M and V, that would not tell us anything necessarily about what it means in a psychological domain.

    Fair enough. But the establishment of the visual dichotomy comes first.

    Next comes the question of asking whether these naturally occurring visual dichotomies have any bearing or significance on psychology. This can be tested a variety of ways, using attention-tests, @teatime has some ideas on this. It can also be tested via manifested behavioral traits, political beliefs, etc.

    Thus, strictly speaking the path CT would take forward into scientific verifiability would take it further away from Jung and typology — but into a realm where it could potentially claim the existence of things like M and V, as natural phenomenon occurring as persistent variations in the population, that correlate with certain psychological/behavioral/attentional traits.

    Whether we wish to say, in retrospect — “This is essentially the Jungian Ne/Si and Ni/Se axis!” — would come later as a commentary. However, that is the situation we’re in. If Jung is ever to be validated, it would be by something like this. Someone will come along and produce experimental evidence for naturally occurring dichotomies that do seem to correlate to an older postulated difference in the population.

    But the new standard, and the new reality of what a “function” means, would now be defined as such (^ in this technical way). So that “what is Ne” would be met with an answer to the effect of “well I don’t know what you mean by Ne, but insofar as something approximating the Jungian “Ne” has been objectively described by the CT system VC-31 (aka “Ne”) is these things: [ insert population data ].



    I’m writing this very fast on the way out the door, so please pardon any typos or errors in general. This is a bit of hypothetical scenario I’m portraying, but necessary to paint a picture of what CT is aiming at, and how it’s treating its relationship to typology.

    Hope that makes sense 🙂

    • This reply was modified 1 year, 3 months ago by Auburn.
    • This reply was modified 1 year, 3 months ago by Auburn.
    • This reply was modified 1 year, 3 months ago by Auburn.
    • This reply was modified 1 year, 3 months ago by Auburn.
    The Doctor
    Participant
    • Type: Unknown
    • Development:
    • Attitude: Unknown

    I understand and agree with the approach you’re taking for internal validity. But, without proof of the existence of the functions themselves, it’s a catch 22. CT makes the assumption of functions as a byproduct of the subjective interpretation of signals.

    At which point, we must ask.. if the functions are real at all, what proof do we have of them?

    And just what is their reality?

    In what way are they real and what do they mean?

    This is the heart of the matter. That’s why I’m suggesting another means of measure to verify this via triangulation. While Nardi’s data isn’t proving the existence of functions, its empirical approach implies their existence, and if his methodology comes to the same conclusion as CT’s methodology with enough correlation accuracy, then both systems could be seen to validate the existence of an abstract concept that cannot otherwise be verified through subjective interpretation of data.

    This could be a big win for both of you and your approaches if they align well enough.

    EpicEntity
    Participant
    • Type: SeTi
    • Development: l--l
    • Attitude: Directive
    I was kind of rifting, but I think I might have inadvertently stubble upon a worth wild question. If it doesn't make sense please disregard.

    If my thoughts about Socionics’s majority conscience are true it has the compacity for people to get along with each others (pseudo-understanding) in living a normal live apart from truth. I don’t know if I can live in a potential lie, yet always searching for the objective truth in it self is taxing. Currently I’m thinking once you just accept something is truth what else happens. Would life present new adventures and new areas of growth in the form of de-stagnation?

    As matter of fact the adventure of diving underneath the layers of established truth. Coming at truth from this perspective reasoning to me seems to be exciting enough to create a live wroth living. You don’t get to the core of the universe but you do get to look it from the outside in of what has been so familiar for so long and perhaps find more universal truths (vectors that point toward core of the universe) as natural consequent. @The Doctor @Auburn Can CT and/or Brain Mapper become a vehicle for such a phenomena?

    <script src=”chrome-extension://hhojmcideegachlhfgfdhailpfhgknjm/web_accessible_resources/index.js”></script>

    • This reply was modified 1 year, 3 months ago by EpicEntity.
    • This reply was modified 1 year, 3 months ago by EpicEntity.
    • This reply was modified 1 year, 3 months ago by EpicEntity.
    Auburn
    Keymaster
    • Type: TiNe
    • Development: l--l
    • Attitude: Adaptive

    Importing a few more…

    Echidna1000

    @Auburn | TiNe These are good GIFs. However, that is a strong step towards test-retest reliability, not towards criterion validity. Correct me if I’m wrong, but the criterion validity would be achieved if you can take your near-perfected process of identifying reoccurring facial/bodily movements and appearances and show that they are reliably associated with aspects of Jungian type that have been measured via independent means, e.g. interviews, surveys, questionnaires, etc. To do this though, you need to have typed a lot of people without using Vultology, but using these more traditional methods of assessing personality. Then you’d need to independently have the same sample record their videos and have CT specialists assess their type through Vultology. That will tell you quickly if Vultology has criterion validity.

    Echidna1000

    @Auburn | TiNe -One question on this: would you say that, putting aside the Vultology part and focusing purely on the CT take on Jungian psychological type theory, would you say your approach to the theory is more robust than Socionics? Is there a similar level or rigour in the definition of Fe, for instance? I ask because if not, I think perhaps the most potentially groundbreaking fusion could be to take Socionics typings, through the more traditional methods, and then identify what shows up as clear patterns from Vultology. It could be what makes ‘VI’ finally work.

    Auburn | TiNe

    @Echidna1000

    Correct me if I’m wrong, but the criterion validity would be achieved if you can take your near-perfected process of identifying reoccurring facial/bodily movements and appearances and show that they are reliably associated with aspects of Jungian type that have been measured via independent means, e.g. interviews, surveys, questionnaires, etc.

    Good question! Well, yes and no. It has to be appreciated that CT is not a “way to prove” Jungian types right — because that will never happen; at least not the way people may think. The Jungian typologies that are existent right now within the typological sphere lack objective proof of their own premises – and thus are a flimsy point of comparison. This presents a methodological problem, like trying to measure a yardstick’s accuracy by using one’s own feet, to see if it really is three feet long. Now this metaphor’s a little off because I’m not suggesting vultology is as flawless as the yardstick yet, but neither is the jungian psychology. However, vultology does have to be measured against a psychological match for it to be anything beyond a body-movement tracking system — but in the terms that it is claiming. For example, in CT a person’s eyes disengaging down (Ji-2) midway through conversation is hypothesized to reflect an internal state of:

    • 1- disengagement from the environment (neglecting)
    • 2- focusing in the precision of an idea by judging its accuracy (i.e. semantics)
    • 3- checking an idea’s alignment to one’s sense of right/wrong

    We can check this by various other means to see if it’s true. We can analyze their language and see if they clarified something (i.e. added disclaimers, made a more keen distinction, etc). We can check if this clarification reflects a desire towards more proper discrimination in data (J), and alignment to oneself (Ji).

    We can check this with 100 people and see if Ji-2 as a visual signal does correlate to the above three psychological points being postulated. And the same sort of experimentation can be done for each of the other signals, or signal groups. If the experiments show a positive correlation, and a high one at that, then… if we see the 101’st person come along and they display high amounts of Ji-2, we can predict that they will have large amounts of points 1-2-3 above, happening cognitively in them. So yes, vultology does need to be checked for psychological validity but not automatically against any other model. That is not the right point of comparison because CT isn’t claiming to be other models — so the alignment or misalignment it may have to them is a separate thing.

     

    Echidna1000

    @Auburn | TiNe – it is entirely fair to build one’s own system and evaluate it on its own merits. Indeed, what you say would be helpful for establishing Vultology as something which does read internal mental processes through external expressions. At the same time, it’s clear from your website that there is significant connection drawn between these visual tells, like J-2, and Jungian cognitive functions, at least in terms of the terminology used. As such, I would say that once you have managed the task you describe and demonstrated the criterion validity of Vultology reading inner mental processes, the next step would be to establish the criterion validity between the prevalence of those mental processes and Jungian psychological type. At the same time, there is work that ought to be done on the purely Jungian side, for instance, demonstrating that psychological asymmetry is actually a thing. Factor analysis could also be helpful, both for determining if information really does get categorised into 8 different kinds, and for determining which patterns of expression are the most salient and appropriate for Vultology.

     

    Auburn | TiNe

    @Echidna1000 Yep! on its own terms. But yes I agree with the need to clarify its alleged involvement with Jungian types. If the CT model can be independently confirmed by more scientific research (and there is a concivable pathway to that, without going through any jungian channels), then the confirmation of such a model would be first and foremost the validation of a vultological-psychological truth of humanity. Secondarily it would be a validation that Jung was onto something.. by virtue of how strongly it parallels many of the concepts postulated by Jung. But the difference here is that CT, if verified as true, would validate what Jung speculated more loosely on. It is not the same as providing support for the Jungian system ‘as such’; as created by the original founder and predecessors. This is because whatever phenomenon is actually there, has to be understood in its own objective terms, not in the terms we have put forth hitherto. In this sense, CT is a very external model and approaches cognition from the outside-in.

    It’s fundamentally about cognition, but the pathway through it is inverted, so rather than starting with clear outlines of information domains, we focus on clusters of visuals and their psychological correlates, and then try to triangulate out a psychological definition from that data.

    Echidna1000

    Thank you for that @Auburn | TiNe, very clear. I would say in contrast that Socionics is much more inside-out.
    • This reply was modified 1 year, 3 months ago by Auburn.
    GreenCoyote
    Participant
    • Type: SeTi
    • Development: lll-
    • Attitude: Adaptive

    Cue Ti absorption. Or should i say ALE absorption 😁

Viewing 8 posts - 1 through 8 (of 8 total)
  • You must be logged in to reply to this topic.

© Copyright 2012-2020 J.E. Sandoval
SEE HERE FOR MORE INFORMATION

DISCLAIMER

The content on this site is not
intended for medical advice, diagnosis,
or treatment. Always seek the advice
of your physician or other qualified
health provider with questions you
may have regarding a medical condition.
For more information visit this link.

SHARE: FACEBOOK, SUPPORT: PATREON